Determinism and the Ability to do Otherwise

Citation preview

Determinism and the Ability to do Otherwise Winston Nesbitt; Stewart Candlish Mind, New Series, Vol. 87, No. 347. (Jul., 1978), pp. 415-420. Stable URL: http://links.jstor.org/sici?sici=0026-4423%28197807%292%3A87%3A347%3C415%3ADATATD%3E2.0.CO%3B2-Y Mind is currently published by Oxford University Press.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/oup.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

The JSTOR Archive is a trusted digital repository providing for long-term preservation and access to leading academic journals and scholarly literature from around the world. The Archive is supported by libraries, scholarly societies, publishers, and foundations. It is an initiative of JSTOR, a not-for-profit organization with a mission to help the scholarly community take advantage of advances in technology. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Fri Jun 22 07:33:24 2007

Determinism and the Ability to do Otherwise W I N S T O N N E S B I T T AND S T E W A R T C A N D L I S H

I n a previous paper,l we criticised the following argument, which we called 'argument A': If a man could not have done otherwise than he in fact did, then he is not responsible for his action. But if determinism is true, it is true of every action that the agent could not have done otherwise. Therefore, if determinism is true, no one is ever responsible for what he does. We noted there that attacks on argument A usually concentrate on its second premiss, and argued that these attacks are superfluous, since the argument either involves equivocation or requires a wildly implausible reading of that second premiss. In the course of our discussion we pointed out that in any case, the appropriate response to the usual reading of that premiss, that determinism (understood as the thesis that all events are caused) is incompatible with one's ability sometimes to do otherwise than one in fact does, would not be to produce elaborate refutations; on the contrary, rather than set about such demonstrations that determinism is compatible with one's having the ability to do other than one does, we should ask why on earth the suggestion of incompatibility should be taken seriously. For while it is clear why it should be thought that if it is determined that one will do something on a given occasion then one will not in fact do anything else on that occasion, it is much less clear why it should be thought that if it is determined that one will do something on a given occasion, then one mill be incapable of doing anything else on that occasion. We shall now try to cast some light on this puzzle; that is, we shall provide what seems to us a plausible explanation of its having been thought that determinism is at least prima facie incompatible with one's having the ability sometimes to do other than one does, and that the onus of proof is therefore on those who wish to deny this incompatibility. This explanation will do something to show that the onus of proof should rather be on those who wish to maintain the incompatibility, and we shall strengthen this conclusion by demonstrating that one of the best known and most generally accepted reasons for believing in the incompatibility turns out to be no reason at all. I

\\'inston Nesbitt and Stewart Candlish, 'On Not Being Able to Do Otherwise', Mind, lxxxii, n.s., no. 327 (July 1973), 321-330.

416

W I N S T O N N E S B I T T A N D STEW'ART C A N D L I S H :

A notorious fallacious argument for scepticism moves from the fallibility of our means of obtaining information about the world to the conclusion that knowledge is impossible. I t runs thus: If you know something, you cannot be wrong. But you can always be wrong. Therefore, you can never know anything. Almost as well known as this argument itself, though not always well understood, are various of its failings. We are interested here in the first premiss, its defects and the sources of its plausibility. The main source is the fact that the premiss seems to be a direct consequence of, perhaps even an alternative way of stating, the necessary truth that knowing that P is incompatible with being wrong about P (i.e. mistakenly believing that P). However, what we may infer from this incompatibility is not that if one knows something one cannot be wrong, but merely that if one knows something one is not wrong. What accounts for the temptation to make the former inference is the seductiveness of the modal fallacy 'Necessarily, if p then q ; therefore if p, then necessarily q'. For from the incompatibility of knowing and being wrong follows the necessity of the conditional 'If X knows that P, then X is not wrong about P'; and one is led by this fallacy to transfer the necessity from the conditional as a whole to its consequent, concluding that if X knows that P, then X cannot be wrong about P (i.e. he is necessarily not wrong). Therein lies the plausibility of the thought that the incompatibility of knowing and being wrong requires that knowing something involves being in a position where mistake is impossible (or, more extremely and less plausibly, where knowing it is logically unavoidable). If anyone thinks that it does follow merely from the incompatibility of knowing and being wrong that if one knows something one cannot be wrong, it can be pointed out to him that he should presumably find conclusive such an argument as, 'If you stand up, you cannot be lying down. But you can always be lying down. Therefore, you can never stand up'. For standing is incompatible with lying down; and it is certainly always logically possible that one is lying down (or, taking the other reading of the ambiguous second premiss, it is logically possible that one is always lying down). Needless to say, to suggest, even correctly, that some assertion owes much, or even all, of its attractiveness to a confusion, is to do nothing towards showing that assertion to be false. But what we have said shows that 'If you know something, you cannot be wrong' is not a mere logical consequence of the indisputable fact that knowing is incompatible with being wrong; so that it is not, as it might seem, obviously true. And failing some argument for its truth, we need not be troubled by any argument of which it forms a premiss.

We can now return to our discussion of the claim that if determinism is true one always lacks the capacity to do other than one does. (Hereafter,

D E T E R M I N I S M AND A B I L I T Y T O D O OTHERTYISE

4'7

'the incompatibility thesis'.) As we have pointed out, the incompatibility thesis is more often asserted or presumed than argued, for it is generally taken that the thesis is prima facie more plausible than its denial. This puzzling, since it is far from obvious why it should be thought that determinism has the peculiar consequence that on any particular occasion, one's abilities are restricted in number to just one, namely the ability to do precisely as one does. T h e explanation of its being thought that the incompatibility thesis needs no argument, we suggest, is the same as that which we offered for its being thought undeniable that if one knows something, one cannot be wrong, that is, that the incompatibility thesis has been taken to be a direct consequence of the genuine inconipatibility between determinism and one's doing other than one in fact does 011 any particular occasion.l We say that this incompatibility is genuine, but it may well be suggested by those too enthusiastically on our side that it is as spurious as its fallacious derivative. For the truth of determinism alone does not entail that one will or will not do any particular thing on any particular occasion. I t will not be denied, however, that if determinism is true, then whatever one does at any given moment, it is determined that one does that and nothing else. And what it means to say that one is determined to do something is that from the statement of determinism (i.e. universal causation) and a true description of the total (or relevant part of the) state of the cosmos (including all causal laws) immediately prior to the action, there follows a statement that that particular action will be done. This means that determinism is after all at least conditionally incompatible with one's doing otherwise than one does at any given moment. And as the condition in question is a description which is ex hypothesi true, we may speak elliptically, though not misleadingly, of a genuine incompatibility between determinism itself and one's doing otherwise than one in fact does.2 Our suggestion, then, is that the incompatibility thesis has appeared self-evident to many philosophers because it seems to be a direct consequence of the fact that determinism is incompatible with one's doing otherwise than one in fact does on any given occasion. For from this incompatibility we can deduce the truth 'Necessarily, if determinism is true, then one will not do otherwise than one does on any particular occasion', just as from the incompatibility of knowing and being wrong I

2

Naturally, clear examples of this confusion are hard to find since a measure of concealment is necessary for such things to be thought, but our diagnosis is at least a highly plausible one of the cases of Roderick Chisholm and Richard Taylor in Hook, S. (ed.) Determinism and Freedom (Collier, New York, 1961). Cf. in particular the second premiss of the argument Chisholm discusses: 'If we make a choice under conditions such that, given those conditions, it is (causally but not logically) impossible for the choice not to be made, then the choice is one we could not have avoided making.' Obviously it is in any case logically impossible to do otherwise than one does at any particular time as well as doing what one in fact does at that time. However, what we suggest determinism to be incompatible with is, of course, not one's achieving this feat, but one's doing otherwise than one does at any given time instead of doing as one in fact does at that time. This distinction disposes of many apparent contradictions and tautologies involved in talk of doing otherwise than one in fact does.

4 1 ~

W I N S T O N NESBITT AND STEWART C A N D L I S H :

we could deduce the truth 'Necessarily, if one knows something, one is not wrong about that thing'. And just as tempting as the move from this latter truth to 'If one knows something, one cannot be wrong about that thing' (it hardly seems a move at all) is the equally fallacious move from the former to 'If determinism is true, then one cannot do other than one does on any particular occasion'.

Our arguments have shown, we hope, that the initial and not inconsiderable plausibility of the incompatibility thesis is spurious, and that the onus of proof should accordingly be on those wishing to assert the thesis rather than on those denying it. However, as we have remarked, this is not where the onus has been thought to lie, for a number of philosophers have made attempts to show that determinism can be reconciled with our being able sometimes to do otherwise than we in fact do; and whether or not such attempts have been successful is at least a moot point. Now it might be thought that the difficulties encountered by such attempts in themselves provide support for the incompatibility thesis-thus Austin, in his celebrated 'Ifs and Cans', suggests that his criticisms of the attempts by Moore and Nowell-Smith to reconcile determinism with 'the things we ordinarily say about what we can do and could have done' go some way towards showing that such a reconciliation is not possib1e.l In view of this, we wish to conclude by arguing briefly that one well known type of attempted reconciliation of determinism and our ability to do otherwise than we do has been given unduly short shrift, since what has been thought to be a fatal objection to it is in fact far from con~lusive.~ T h e objection we have in mind is made by Austin himself in the paper just mentioned. H e discusses the suggestion that 'A could have done X' is to be analysed as a hypothetical statement to the effect that A would have done X if some condition, such as his having chosen, or wanted, to do X, which was not in fact fulfilled, had been fulfilled. If accepted, this analysis would show that determinism is after all compatible with its sometimes being true of someone that he could have done otherwise than he did; for determinism is certainly compatible with its sometimes being true of someone that he would have done otherwise than he did if some condition, which was not in fact fulfilled, had been fulfilled. T h e version of this analysis which Austin finds most plausible is the suggestion that 'He could have done X' means 'He would have done X, if he had tried'; but he presents an example which, be believes, shows that it has 'no more' than plausibility. Briefly, the example is that of a golfer who I J. L. Austin, 'Ifs and Cans', in Austin, Philosophical Papers (Oxford 2

University Press, 1961), 179. This is in fact a slight exaggeration of the true situation, which seems to be as follows. Philosophers have been so impressed with Austin's detailed argument against certain analyses of 'could have' that they have tended to assume that there is something unsatisfactory about the general form of such analyses. But it should be noted that the brief argument outlined in the following paragraph is Austin's only direct argument against the particular analysis we think to be a strong candidate for the correct one, as opposed to arguments against related analyses and refutations of arguments in favour of analyses.

D E T E R M I N I S M AND A B I L I T Y T O DO O T H E R W I S E

4I9

misses an easy putt and kicks himself because he could have holed it. Here the agent could have holed the putt, but it is false that he would have holed it had he tried to, for he did try to hole it, and failed. Hence, it cannot be that 'He could have done X' means 'He would have done X, if he had tried' (ibid. p. 166, n.). Austin's example might be objected to on the ground that it begs the question by assuming the intelligibility of its being true of someone both that he could have done something and that he would not have succeeded in doing that thing if he had tried. However, to take this line would be to commit oneself to holding that one does not have the ability to do something unless one invariably succeeds when one tries to do that thing, and Austin seems right to us when he suggests that our ordinary understanding of the word 'can' allows that '. . . a human ability or power or capacity is inherently liable not to produce success, on occasion, and that for no reason' (loc. cit.). We accept, therefore, that he has shown that 'He could have done otherwise' and 'He would have done otherwise, if he had tried' do not have the same meaning. Now if the issue were merely that of whether or not the analysis under consideration is correct, this would be the end of the matter. But those who put forward this form of analysis have, as Austin was aware, done so in the course of attempting to reconcile determinism with one's having the ability sometimes to do otherwise than one does. I t is therefore very much to the point to ask, not merely whether Austin has refuted the suggested analysis, but also, how serious is the damage done by his refutation to the enterprise in the course of which such an analysis has been offered-namely, the enterprise of showing that determinism is consistent with the common belief that we can sometimes do otherwise than we in fact do; and the answer to this latter question is, as we shall try to show, 'Not very serious at all'. Austin has shown, as we concede, that 'He can do otherwise' does not entail 'He will do otherwise if he tries'. However, for all that is shown by his example, it can still be claimed that the reverse relationship holdsthat 'He will do otherwise if he tries' entails 'He can do otherwise'. (And similarly for the different tenses of the two sentences.) This would not be enough to save the claim that the two sentences have the same meaning; but it would be enough to save the claim that determinism is consistent u

with our having the ability sometimes to do otherwise than we do. For as we noted earlier, determinism is quite compatible with its sometimes being true of someone that he would have done otherwise if some condition, such as his having tried to do otherwise, which was not in fact fulfilled, had been fulfilled; and thus, if 'He would have done otherwise if he had tried' entails 'He could have done otherwise', then neither is determinism incompatible with its sometimes being true of a man that he could have done otherwise. Of course, it may not be true that 'He would have done otherwise if he had tried' entails 'He could have done otherwise'. But as we have pointed out, Austin's example in no way shows that it is not true; and moreover, it is highly plausible to suggest that it is in fact true, since to say 'I would have done so if I had tried, but nevertheless I could not have done so' is to give a strong impression of contradicting oneself.

NESBITT AND CANDLISH

: DETERMINISM

v What exactly have we argued here concerning the incompatibility thesis? We have not attempted to argue that there are no grounds at all for believing it to be true. We do claim, however, that we have exposed at least a major source of its plausibility, and have shown this source to reside in a common modal fallacy. We have also shown that the common belief that Austin has refuted the most favoured argument against the thesis is in fact groundless, and our paper has, therefore, provided some reason for thinking that the incompatibility between determinism and one's ability to do otherwise is only apparent. It remains to trace the consequences of all this for what, at the start of this paper, we called 'argument A'. In our previous discussion of this argument (see p. 415, n. I), we left open the possibility of rewriting it in a way which would make it consistent (though it would then have implications much more dire than those envisaged by its proponents) and thus immune to the criticisms we had so far brought against it (op. cit. p. 330, note). But the second premiss of A is a statement of the incompatibility thesis, and what we have said here on that subject shows that even the best possible version of A is not only extreme in its implications, but still inherently dubious. Argument A still has a powerful appeal to the imaginations of philosophers who think about freewill and determinism, but it appears more and more that that appeal is entirely superficial. U N I V E R S I T Y O F WESTERN AUSTRALIA